Early diagnosis of systemic infection in humans and prediction of clinical outcome both remain elusive, despite the critical need for both. The two basic approaches for diagnosis and prognosis, direct identification of the infectious agent and measurement of host response features, have been limited by reliance on cultivation, pre-test assumptions about specific agents, limited breadth of coverage, and an insufficient signal-to-noise ratio. Recent advances in DNA sequencing technology and computational tools for data analysis provide new opportunities for characterizing complex mixtures of DNA and RNA molecules in a minimally-biased fashion, and for detecting rare sequences of interest in clinical specimens. The goals of this project are to characterize microbial and viral genomic sequences in the peripheral blood of healthy subjects, as well as subjects with fever (and in some cases, known causative infectious agents) at the time of clinical presentation, to characterize human genome-wide RNA patterns in peripheral blood, and then to integrate these complementary forms of data so as to derive putative diagnostic and prognostic signatures for systemic infection. We propose three aims; (1) characterize microbial and viral sequences and host RNAs in peripheral blood of non-febrile, healthy subjects; (2) characterize microbial and viral sequences and host RNAs in human blood during febrile illness; and (3) identify and validate diagnostic and prognostic biomarkers during early stages of febrile illness based on microbial and viral sequences, host RNAs, and clinical features. Our long-term objectives are to validate novel early diagnostic and prognostic signatures for several classes and species of pathogens and their related clinical syndromes, with attention to early predictors of subsequent critical illness requiring intensive care. Sequence-based signatures will serve as the basis for the design of probes on CMOS electrochemical-based active microarrays and other technologies, so that they can be exploited and deployed at the point-of-care in an easy-to-use format.